Prediction models for the development of dental caries have been investigated by many researchers. To date, the variables shown most ofen to be correlated with the development of dental caries are the subject's DMFS prevalence and the number of permanent surfaces that exist in the dentition. In a previous study, Kingman has shown that the incorporation of the various surface susceptibility levels for caries into the subject's caries prevalence assessment can improve models used for caries prediction. The current investigation focuses on differences in surface susceptibility level to caries attack and also the separate counts of Dm F and M surfaces diagnosed for the subject. A weighted caries experience is derived using these factors and tested for its predictive power, using recently conducted clinical trial data sets.